Evaluating a parent training program: Scale analysis and the effects of Systematic Training for Effective Parenting (STEP) on child and parent behavior

Two studies examined archival pretest and posttest data collected from 490 parents who participated in a 7-week Systematic Training for Effective Parenting (STEP; Dinkmeyer & McKay, 1976) program in a large, suburban, public school district in Rhode Island between 1994 and 1999. The first study cross-validated the underlying structure of selected subscales from the Behavior Rating Profile, Second Edition (BRP-2; Brown & Hammill, 1990) and STEP Parenting Assessment Technique (SPAT; Dinkmeyer, 1981). Principal components analyses revealed two components (Externalizing Behavior and Internalizing Behavior) with good to very good coefficient alphas on a revised version of the BRP Parent Rating Scale, and one component (Negative Parenting Behavior) with a comparatively weaker coefficient alpha on the SPAT Parenting Behavior Scale. Thus, only the BRP Parent Rating Scale Revised showed internal consistency reliability with this sample of parents. Test-retest stability was examined for 230 parents who completed both the pretest and posttest measures. All Pearson product-moment correlation coefficients calculated at each of three STEP levels (early childhood, elementary, and teen) were statistically significant (p \u3c .01), indicating that parents\u27 assessment of their children\u27s and their own behavior were relatively consistent over a 7-week period of time. The second study explored the effects of parent training on parents\u27 perceptions of their target child\u27s behavior and their own behavior. A 2 x 3 repeated measures multivariate analysis of variance for time and STEP level found a significant overall improvement in scores from pretest to posttest at all three age groups (p \u3c .01). Follow-up univariate tests determined that Externalizing Behavior, Internalizing Behavior, and Negative Parenting Behavior measures all improved significantly from pretest to posttest. For STEP level, post-hoc comparisons revealed significant mean differences between younger and older children for Externalizing Behavior and Internalizing Behavior. The results indicate that the BRP Parent Rating Scale Revised (which included three items authored by staff members in the sampled school district), demonstrated strong psychometric properties. However, the SPAT Parenting Behavior Scale (the measure developed to assess the STEP program) was not a psychometrically sound instrument. While the STEP program appeared to improve child and parent behavior, causal conclusions could not be drawn from the present design